Modern recovery systems for network enabled digital devices are substantially dependent on operating system functionality. As such, the operating system is often responsible for monitoring and responding to various operational conditions, such as operating system panic, failure, and/or lock-up. Unfortunately, this results in both a condition and a monitoring mechanism being substantially dependent on the same operating system. As a crippling operational condition may be situated within the path of observed behavior, any resulting recovery data obtained by the monitoring mechanism may also suffer from the same poisoned data path or condition that originally crippled the operating system.
In an effort to partially address these difficulties, some operating systems, such as the Windows® operating systems, define a special path for a disk driver of the operating system to log failure information to storage. Unfortunately, even having a special fast path does not eliminate the potential problems relative to a poisoned data path or a crippling operational condition, because the configuration still suffers from the possibility that the failure may be part of the operating system kernel, on which retrieval of the logged information depends.